News/Event Training Coverage: 2026-07-13

Run news_training_coverage_20260720T054509Z | Generated 2026-07-20T05:47:08.255957+00:00

News/Event Training Coverage: 2026-07-13

Run: news_training_coverage_20260720T054509Z

This report tracks whether the news/event data lake is filling at the raw API, normalized evidence, and matured outcome layers.

Summary

News Model Use Policy

This is a report-only guardrail. It translates source quality, classification quality, duplicate pressure, and matured forecast outcomes into how news/event evidence may be used by research models.

Policy AreaStatusObservedRequired/ContextGap/Review RowsRatioModel Usage PolicyOperator Action
factual_event_training_evidenceresearch_feature_candidate563675168353411198590.3348research_feature_candidate_not_promotion_permissionUse factual/mixed event rows as research features only after source, entity, confidence, and duplicate gates are applied.
classification_quality_gateclassification_review_required125952316835344240110.7481downweight_or_exclude_review_bucketsDownweight or exclude buckets with low source quality, low entity relevance, low confidence, or unknown publisher class before model weighting.
duplicate_cluster_handlingdedupe_required_before_weighting489281683534489280.0291dedupe_cluster_cap_before_weightingCap repeated article clusters so one syndication wave cannot dominate an event signal.
publisher_source_quality_gatesource_quality_usable_with_reviews149988016835341836540.8909source_weighting_candidate_with_unknowns_reviewedKeep publisher/source-class weights visible; unknown or low-quality publisher classes need review before gaining weight.
opinion_forecast_track_recordforecast_track_record_research_candidate3336202111985901.0000track_record_overlay_only_not_factual_truthTreat opinion/forecast rows as reliability-overlay candidates only; do not mix them into factual event truth.
aggregate_news_signal_policynews_signal_research_only_quality_review125952316835346076650.7481research_only_until_lift_and_promotion_gates_passUse the news signal in research/ablation only until lift, leakage, stability, and promotion gates pass.

Raw Coverage By Plan

PlanCoverage RowsSymbolsMin StartMax EndSucceededEmptyFailedRaw Bytes
benzinga_newsfeed_daily6005192026-07-022026-07-161599028695
benzinga_newsfeed_recent_backfill27124242023-01-012026-07-15902622013111534
fmp_grades_consensus_research_pool1316613702026-06-032026-07-161286929701815904
fmp_grades_historical_research_pool225011572016-06-052026-06-2022464037880578
fmp_grades_research_pool1316613702026-06-032026-07-16128433230381606271
fmp_press_releases_recent_backfill31867892026-01-102026-07-1320571129011280067
fmp_press_releases_research_pool1316613702026-06-032026-07-16130491170131998638
fmp_price_target_consensus_research_pool1316613702026-06-032026-07-161240576101590660
fmp_price_target_summary_research_pool1316613702026-06-032026-07-161240576104285554
fmp_stock_news_historical_backfill22355921142011-01-012026-07-138671713684201543138930
fmp_stock_news_research_pool1241613932026-06-022026-07-02122941220304233577

Source Family Coverage

This rollup groups raw coverage plans into provider/source families so Benzinga, FMP stock news, issuer press releases, and analyst-action feeds can be reviewed separately before increasing source weight or backfill volume.

Source FamilyPlansCoverage RowsSymbolsMin StartMax EndSucceededEmptyFailedRaw BytesOperator Action
FMP stock newsfmp_stock_news_historical_backfill, fmp_stock_news_research_pool23597535072011-01-012026-07-139901113696401847372507coverage available; monitor source quality and outcome maturity
FMP analyst actionsfmp_grades_consensus_research_pool, fmp_grades_historical_research_pool, fmp_grades_research_pool, fmp_price_target_consensus_research_pool, fmp_price_target_summary_research_pool5491466372016-06-052026-07-165276821460427178967use as forecast/opinion evidence and keep track-record calibration visible
FMP press releasesfmp_press_releases_recent_backfill, fmp_press_releases_research_pool1635221592026-01-102026-07-161510612460143278705use as issuer event evidence; watch duplicate clusters before weighting
Benzinga market newsbenzinga_newsfeed_daily, benzinga_newsfeed_recent_backfill33129432023-01-012026-07-16913221013140229validate Benzinga event taxonomy/source quality against FMP and matured outcomes

News History Expansion Priorities

This derived queue ranks the next bounded news-history work from local coverage and quality evidence. It is guidance only: it does not call providers or queue tasks.

PriorityExpansion AreaStatusObservedRequired/TargetGap/ReviewSuggested ScopeOperator Action
1benzinga_licensed_market_news_validationbenzinga_validation_seed_sparse91500409targeted_top_candidate_windows_not_broad_empty_retriesUse bounded Benzinga windows on current opportunity, stop-loss, and sector-exposure candidates; compare taxonomy/source quality before increasing source weight.
2fmp_stock_news_historical_depthfmp_historical_stock_news_depth_gap99011100000989window_first_historical_stock_news_monthsContinue window-first historical stock-news backfill for sparse or underfilled years before asking for new paid news sources.
4sparse_backfill_month_reviewsparse_backfill_months_present606review_months_before_repeating_provider_callsReview sparse months before rerunning them, so known-empty windows do not consume daily/provider capacity.
6classification_and_duplicate_review_pressureclassification_review_pressure_present4729390472939review_quality_buckets_before_source_weight_increaseClean or downweight review-heavy buckets before giving new backfilled news more model influence.
6factual_event_depthfactual_event_depth_ready5636755050600issuer_press_release_trusted_market_reporting_sec_event_sourcesFactual/mixed event share is adequate for research; keep classification and duplicate gates active.
8normalized_event_month_depthnormalized_event_month_depth_ready18180extend_training_history_months_after_raw_backfillNormalized event months are broad enough for this report window; use outcome and quality gates to steer expansion.

Stock-News Backfill Months

MonthWindowsSymbolsSucceededRaw Bytes
2026-07-011017895184310904
2026-06-0111989838038009859
2026-05-0144722031256228841208
2026-04-0132132031203133755713
2026-03-0122861174116520693705
2026-02-0111783251256658
2026-01-0144591174232442584017
2025-12-01106710076382465268
2025-11-0113951011100015915833
2025-10-0122501159116620798351
2025-09-0122211159114714471278
2025-08-0125301170105011581963
2025-07-0129091173112516592912
2025-06-0121521152112115175250
2025-05-0143041152213029512223
2025-04-0121521152112816620512
2025-03-0121521152105312524659
2025-01-0143041152220532969473

Normalized Event Months

MonthEvent RowsSymbolsOpinion RowsFactual/Mixed Rows
2026-07-017099158542312868
2026-06-012596520991114814817
2026-05-013706722191928517782
2026-04-013727921251917918100
2026-03-012475913931208012679
2026-02-013160113411897412627
2026-01-012787513341691110964
2025-12-01205251259113749151
2025-11-01241191297141369983
2025-10-01283241375192309094
2025-09-01172231262106946529
2025-08-01223921348148257567
2025-07-01259021330183647538
2025-06-0115212125084506762
2025-05-01267371300176409097
2025-04-01294961284203849112
2025-03-01181331262105407593
2025-02-01262781286168189460

Event Taxonomy And Source Quality

Event KeyFamilyKindPublisher ClassRowsSymbolsLow SourceLow EntityLow ConfidenceUnknown PublisherTrue Duplicate ClustersExtra Duplicate ArticlesAvg ConfidenceAvg Source QualityAvg Entity Relevance
analyst_revisionanalyst_revisionopinion_forecastanalyst_forecaster10715721362007621804866010171520.55920.62001.0000
issuer_releaseissuer_releasefactual_eventissuer_wire_release19173922190120000000.74970.70000.9919
legal_regulatoryregulatoryfactual_eventissuer_wire_release564891331020600000.74590.70000.9948
product_contract_partnershipproduct_contractfactual_eventissuer_wire_release408071522028100000.74710.70000.9914
earnings_resultearningsfactual_eventmarket_analysis32204205032204110321940000.52050.56000.9940
market_opinionmarket_opinionopinion_forecastmarket_analysis268021704268029919610000.61790.56000.9798
corporate_actioncorporate_actionfactual_eventmarket_analysis25384152625384173253490000.52170.56000.9927
legal_regulatoryregulatoryfactual_eventmarket_analysis22921147122921233223030000.52240.56000.9912
capital_returncapital_returnfactual_eventmarket_analysis20511118320511426203410000.52160.56000.9879
financing_liquidityfinancing_liquidityfactual_eventissuer_wire_release179331258017500000.74670.70000.9923
capital_returncapital_returnfactual_eventissuer_wire_release17765135904200000.74530.70000.9948
corporate_actioncorporate_actionfactual_eventissuer_wire_release171381225014600000.74470.70000.9916
cybersecuritycybersecurityfactual_eventissuer_wire_release1572277507800000.74390.70000.9932
product_contract_partnershipproduct_contractfactual_eventmarket_analysis12494112012494348123880000.51960.56000.9815
issuer_releaseissuer_releasefactual_eventmarket_analysis11933123511933299118560000.51940.56000.9801
analyst_revisionanalyst_revisionopinion_forecastmarket_analysis97591375975910051250000.56210.56000.9911
labor_restructuringoperationsfactual_eventissuer_wire_release778573703800000.74470.70000.9930
macro_sectormacro_sectormixed_contextmarket_analysis6402915640246263360000.51660.56000.9683
legal_regulatoryregulatoryfactual_eventtrusted_market_reporting6244280025700000.69850.82000.9800
macro_sectormacro_sectormixed_contextissuer_wire_release508452302100000.74980.70000.9959

Classification Review Queue

This queue is derived from the event taxonomy aggregation above. It ranks event/publisher buckets where source quality, confidence, entity matching, unknown publisher classes, or true repeated article clusters should be reviewed before giving the signal more model weight. A populated cluster hash with only one article is not duplicate pressure.

Event KeyPublisher ClassKindReview RowsLow SourceLow EntityLow ConfidenceUnknown PublisherTrue Duplicate ClustersBucket RowsOperator Action
analyst_revisionanalyst_forecasteropinion_forecast7621800762180486601071572tighten event-key terms or confidence thresholds; dedupe repeated article clusters before weighting
earnings_resultmarket_analysisfactual_event6450832204110321940032204review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
corporate_actionmarket_analysisfactual_event5090625384173253490025384review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
legal_regulatorymarket_analysisfactual_event4545722921233223030022921review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
capital_returnmarket_analysisfactual_event4127820511426203410020511review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
market_opinionmarket_analysisopinion_forecast28754268029919610026802review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
product_contract_partnershipmarket_analysisfactual_event2523012494348123880012494review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
issuer_releasemarket_analysisfactual_event2408811933299118560011933review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
analyst_revisionmarket_analysisopinion_forecast1498497591005125009759review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
macro_sectormarket_analysismixed_context1320064024626336006402review source scoring; tighten event-key terms or confidence thresholds; inspect symbol/entity matching
issuer_releaseissuer_wire_releasefactual_event120001200000191739inspect symbol/entity matching
product_contract_partnershipissuer_wire_releasefactual_event281028100040807inspect symbol/entity matching

Publisher Class Quality

Publisher ClassRowsSymbolsForecast RowsForecast OriginsAvg Source QualityAvg Entity RelevanceAvg Confidence
analyst_forecaster1076529136210765293050.62001.00000.5587
issuer_wire_release3746122242080.70000.99260.7478
market_analysis180565221436561430.56000.98680.5373
trusted_market_reporting289967562930240.82000.98520.6956
licensed_market_news1974312973751580.68000.98080.6623
unknown259157101960.45000.98430.5896
video_or_social_distribution4981738890.35000.94490.5506

Matured Forecast Outcomes

HorizonRowsSymbolsMin PublishedMax ActualHit RateSigned Capture
1111487313932016-07-15 00:00:00+00:002026-07-170.50840.0007
5111413813932016-07-15 00:00:00+00:002026-07-170.52230.0028
20110719113912016-07-15 00:00:00+00:002026-07-170.53640.0102

Forecaster Track Record Readiness

Sample StatusHorizonOrigin RowsMatured RowsEligible RowsAvg Hit RateAvg Signed Capture
review1653858520.5018-0.0001
thin_sample118116811600.4812-0.0006
trainable119211093341850.50500.0006
review5653858520.4906-0.0003
thin_sample517816781570.4255-0.0020
trainable519211086021850.51980.0017
review20653836520.48030.0013
thin_sample2017116601510.4886-0.0044
trainable2019211016951850.52080.0055

Largest Eligible Forecaster Samples

OriginLevelOrganizationAuthorHorizonEventMatured RowsSymbolsHit RateSigned CaptureSample StatusEligible
fmp_analyst_aggregationorganizationFMP analyst aggregationn/a1analyst_revision9377013610.51600.0013trainableTrue
fmp_analyst_aggregationorganizationFMP analyst aggregationn/a5analyst_revision9352913510.53300.0051trainableTrue
fmp_analyst_aggregationorganizationFMP analyst aggregationn/a20analyst_revision9237011760.54040.0141trainableTrue
morgan_stanleyorganizationMorgan Stanleyn/a5analyst_revision786676130.51410.0019trainableTrue
morgan_stanleyorganizationMorgan Stanleyn/a1analyst_revision786676130.50250.0002trainableTrue
morgan_stanleyorganizationMorgan Stanleyn/a20analyst_revision785686130.54520.0097trainableTrue
barclaysorganizationBarclaysn/a1analyst_revision536026560.51770.0011trainableTrue
barclaysorganizationBarclaysn/a5analyst_revision536006560.52690.0028trainableTrue
barclaysorganizationBarclaysn/a20analyst_revision535486560.54050.0092trainableTrue
wells_fargoorganizationWells Fargon/a1analyst_revision479596120.50680.0008trainableTrue
wells_fargoorganizationWells Fargon/a5analyst_revision479566120.53210.0045trainableTrue
wells_fargoorganizationWells Fargon/a20analyst_revision478926110.54280.0118trainableTrue
citigrouporganizationCitigroupn/a1analyst_revision442376370.51090.0010trainableTrue
citigrouporganizationCitigroupn/a5analyst_revision442366370.52640.0027trainableTrue
citigrouporganizationCitigroupn/a20analyst_revision441476360.54260.0115trainableTrue

Active/Queued News Tasks

PlanStatusTasks
No active news tasks

Operator Commands

.\scripts\run-news-event-historical-backfill.ps1 -PoolSize 50 -StartDate 2024-01-01 -EndDate 2024-12-31 -MaxPages 2 -MaxTasks 100 -BatchOrder window-first -PageLimit 250 -Workers 24
.\scripts\run-news-event-historical-backfill.ps1 -PoolSize 1000 -StartDate 2014-01-01 -EndDate 2018-12-31 -WindowDays 365 -MaxPages 2 -MaxTasks 10000 -BatchOrder window-first -PageLimit 250 -Workers 24
.\scripts\run-stock-news-event-intelligence.ps1 -RunKind manual -TrainingHistoryDays 3650
.\scripts\run-news-forecast-track-record.ps1
.\scripts\get-news-training-coverage.ps1